Portal de Periódicos da CAPES

The visual motion aftereffect from mental imagery depends on speed.

2013; Wiley; Volume: 35; Issue: 35 Linguagem: Inglês

1551-6709

Alexia Toskos Dils, Lera Boroditsky,

Visual Attention and Saliency Detection

The visual motion aftereffect from mental imagery depends on speed Alexia Toskos Dils (atoskos@stanford.edu) Lera Boroditsky (lera@stanford.edu) Stanford University, Department of Psychology Jordan Hall, 450 Serra Mall, Building 420, Stanford, CA 94305 USA Abstract When we imagine a train snaking through a desert, does information about the train’s speed make it into our visual mental image? In this paper, we make use of the motion aftereffect illusion (MAE) to test whether the speed of imagined visual motion modulates transfer of adaptation to a subsequent visual motion discrimination task. We compared the effects of viewing slow, medium, and fast motion on the magnitude of the MAE (Experiment 1) with the effects of simply imagining the same motion stimuli (Experiment 2). In Experiment 1 we found that increasing the speed of real visual motion from slow to medium produced a corresponding increase in the magnitude of the MAE, but increasing speed from medium to fast did not. Likewise, imagining slow motion produced a smaller MAE than did imagining medium motion, but the effect leveled of between medium and fast motion. These findings suggest that our mental imagery of motion is specific to the speed of the moving objects, and highlight areas of overlap between mental imagery and visual perception. Keywords: Mental imagery; Motion aftereffect; Embodiment Background When we imagine a car racing by, how visual is the process of creating the mental image? Do the representations we generate include information about how fast the car appears to be going? Or are they invariant to this property of visual motion perception? In this paper, we make use of the motion aftereffect illusion (MAE) to test whether the speed of imagined visual motion modulates transfer of adaptation to a subsequent visual motion discrimination task. Researchers have long debated just how similar imagining a visual scene is to actually witnessing it (Kosslyn, 1981; Pylyshyn, 1973). Previous work examining the metric properties of imagined static scenes has found that information about size (Kosslyn, 1975), distance (Kosslyn, Ball, & Reiser, 1978), and structure (Kosslyn, 1973) is indeed persevered in mental imagery. For example, Kosslyn et al (1978) found that the distance between objects in a mental image is proportional to the physical distance between their real-world counterparts. Participants in their study memorized a fictional map containing several landmarks, and were later asked to “scan” between pairs of landmarks in their mental image of the map. Results showed that the greater the distance between two landmarks on the physical map, the longer it took people to mentally scan between them. Other work has shown that people are capable of mentally performing metric transformations on images of static objects (Finke, Pinker, & Farah, 1989; Shepard & Metzler, 1971). In a study by Shepard and Metzler (1971), participants judged whether pairs of geometric objects were identical to one another or mirror reversed. The authors reasoned that if people solved this task by mentally rotating one object until it aligned with the other, their reaction time should depend on the physical angular disparity between objects. Indeed, participants took longer to mentally rotate objects that would take longer to physically rotate, and vice versa. The metric properties of mental imagery for dynamic scenes have not been studied as widely as for static scenes. One feature of visual motion that has been found to make it into mental imagery is motion direction. Winawer, Huk, & Boroditsky (2008) demonstrated that imagining visual motion in a particular direction is sufficient to produce direction-selective adaptation in the visual system (i.e., produce a visual motion aftereffect illusion). After imagining upward motion, participants were more likely to see a subsequent dynamic stimulus as moving downward, and vice versa. Transfer of adaptation from mental imagery to perception suggests that a common neural mechanism underlies both processes. However, the degree of adaptation from mental imagery was considerably weaker compared to that from real visual motion perception, which sets a limit on the overlap between these two processes. The adaptation paradigm used by Winawer and colleagues provides a unique testing ground for discovering other motion properties preserved in dynamic mental images. In this paper, we ask whether the magnitude of the visual motion aftereffect from mental imagery depends on the speed of imagined motion. If so, does motion speed modulate the MAE from imagery in the same way as speed modulates the MAE from real visual motion perception? That is, is speed yet another feature common to both mental imagery and perception, or is it an area in which internally- generated motion representations abstract away from their externally-generated counterparts? To test these questions, we first measured the effect of speed on the MAE from real visual motion (Experiment 1), and compared that with the MAE from imagining the very same motion stimuli (Experiment 2). In Experiment 1, subjects viewed videos of moving stripes (upward or downward) in three within-subject conditions: slow, medium, and fast. Following each video, participants indicated the direction in which a set of dynamic dots appeared to move. We found that increasing the speed of